Political Futures Market
There has been quite a bit of excitement in the punditocracy of late about the use of “futures markets” as predictors of electoral outcomes. The logic is quite compelling, as Jim Garven explains. While politics is different from commodities, the same dynamic would seem to apply: those with a financial stake in the thing have much more incentive to study the underlying trends and variables and should therefore be ahead of the curve.
Many who hope President Bush is re-elected have taken comfort from the fact that, although Kerry has been slightly ahead in most of the public opinion polls, Bush is comfortably ahead in the market surveys. For example, Pejman Yousefzadeh has an interesting analysis of the state-by-states races that shows “enough to give the President 266 electoral votes–4 shy of the magic number needed to win.” Others (including another piece I read somewhere today but can’t recall) have found similar results.
Nathan Littlefield looks at the actual numbers in the current Atlantic Monthly and points out that “political markets aren’t much better than pundits.” He compares the predicted numbers on July 1 for the last four presidential contests with the electoral outcomes:
2004 July 1 Result
George W. Bush 52.8% TBD
John Kerry 46.0% TBD
2000 July 1 Result
George W. Bush 48.5% 49.7%
Al Gore 49.3% 50.3%
1996 July 1 Result
Bill Clinton 51.9% 54.7%
Bob Dole 43.5% 45.3%
1992 July 1 Result
Bill Clinton 28.0% 43.3%
George H.W. Bush 35.9% 37.7%
H. Ross Perot 35.0% 19.0%
1988 July 1 Result
George H.W. Bush 46.4% 53.9%
Michael Dukakis 50.4% 46.1%
Indeed, the “market” was wildly wrong in 1988 and 1992. It came much closer in the last two races, which is indication of either a fluke or some methodological improvement in the process. My money, as it were, is on the former. It may be true that those who speculate in orange juice futures make better weather forecasters than the National Weather Service. As complicated as the weather is–what with the actions of butterflies in China being so influential–predicting the behavior of 290-odd million Americans is exponentially more so.
If you’re interested in playing, Bush is selling for $470 and Kerry for $431 at the moment.
